02 intel v_tune_session_02

Code Optimization and Performance Tuning Using Intel VTune
Installing Windows XP Professional Using Attended Installation

Objectives

In this session, you will learn to:
Identify time and space complexity
Optimize programming constructs
Identify performance libraries for optimization

Ver. 1.0 Slide 1 of 24


Examining Algorithms

An algorithm is a set of finite steps that accomplish a
specific task.
The efficiency of an algorithm is measured in terms of the
processor time and the memory space that the algorithm
utilizes.
It is necessary to tune the algorithm to make optimal use of
available resources, such as processor time and memory.
To analyze algorithmic complexity, it is important to
understand the time and space aspects of complexity.



Identifying Time Complexity

The time taken to compute the steps involved in an
algorithm is called the time complexity of that algorithm.
The time taken to execute all the steps for a worst-case
scenario is the time complexity of the entire algorithm.
The execution of steps in an algorithm depends on the
conditions specified in it. As a result, the time complexity of
an algorithm is measurable with reference to the conditions
involved in the algorithm.



Identifying Time Complexity (Contd.)

Computing the nth number of the fibonacci series:
Step 1: start
Step 2: input the value of n
Step 3: if (n < = 1) then go to step 14
Step 4: x = 0
Step 5: y = 1
Step 6: write (x + “ “ + y)
Step 7: for (i = 0 to n-1)
Step 8: {f = y + x
Step 9: x = y
Step 10: y = f
Step 11: i = i + 1
Step 12: write (f) }
Step 13: go to step 15
Step 14: write (n)
Step 15: stop




Based on the value of n, two cases can exist in the
preceding algorithm.
If the value of n is less than or equal to 1, time complexity is
constant and does not depend on input, which is n, as
shown in the following table.

Statements Frequency of Execution
Step 2: input the value of n 1
Step 3: if (n < = 1) then go to step 14 1

Step 14: write (n) 1
Total Number of instructions executed 3




If the value of n is greater than 1, time complexity is 4n-2.
This is shown in the following table.
Statements Frequency of Execution

Step 2: input the value of n 1

Step 3: if (n < = 1) then go to step 13 1

Step 4: x = 0 1

Step 5: y = 1 1

Step 6: for i= 2 to n-1 repeat steps 7 to 10 1

Step 7: f = y + x n-2

Step 8: x = y n-2

Step 9: y = f n-2

Step 10: i = i + 1 n-2

Step 11: write (f) 1

Total number of instructions executed 4n-2



Just a minute

What do you mean by the time complexity of an algorithm?

Answer:
The time taken to compute the steps involved in an algorithm
is called the time complexity of that algorithm.



Identifying Space Complexity

Space complexity is the amount of memory that a program
requires to accomplish a task.
Space complexity is a criterion to measure the efficiency of
a program.
For example, consider the following expression:
Return [a + b + b * c + (a +b – c) / (a + b) + 4.0]
If variables b and c are each of 2 bits, the storage requirements
for the preceding instruction will be constant.
If the values of a, b, and c are taken from user input, the
storage space required will vary.



Identifying Space Complexity (Contd.)

Refer to the following algorithm:
Step 1: start
Step 2: function Add (k, m)
Step 3: l = 0
Step 4: for j= 1 to m repeat step 5
Step 5: l = l + k[j]
Step 6: return the value of l
Step 7: stop

For the preceding algorithm:
The space taken by the k array is m units.
The other variables, m, j, and l will take only one unit of space
each.
As a result, you can obtain the space complexity of this
algorithm by adding the space utilized by all the variables.



Examining Programming Constructs

An application developed in any programming language,
such as C, C++, C#, or Java, is based on an algorithm.
Each algorithm consists of several programming constructs,
such as loops, decisions, and functions.
The performance of an application is related to the
programming constructs used in the application.
To achieve the desired level of optimization, it is important
to examine the loops, branching statements, and function
calls used in the program.



Examining Loops

Loops can be optimized by:
► Removing unwanted parts of loops In this technique, you need to
first identify the
► Combining loops When multiple loops work on
decision-making steps present
the same variables, you may
► Using unrolling inside a loop. to the
Unrolling refers
combine the loops.
Aftervaluesbreaking compact
process of
► Reducing work inside loops The identifying these steps,
of some
This helpssimpler statements.
loops into reduce
you need to decide whether the
expressions, time because
computation variables, or
► Using sentinel values loop affects the steps in any
A sentinel do notis a value
constants value of
the total number change
way.is the loop and end of a
that placed at the
► Looking at the order of loops inside improve the
You can
instructions executed
search range. not affect
If the loop of a consumeby the
reduces. does
unnecessarily program
efficiency
► Looking at operators steps, remove prevents you
Sentinel value the unaffected
In terms oftime.
processor efficiency,
changing the order of loops.
stepsperforming theapproach
from from a better additional
Therefore, the loop.
operations such as
task of checkingoutside the
is to bring them the end of
multiplication and division are
the input search than
more expensive string.
loop.
operations such as addition.
You should try to convert all
expensive operations with
cheaper ones.



Examining Functions

As the frequency of calling a function increases, program
execution time increases.
Examining functions is important for performance
optimization.
Functions can be optimized by:
► Using faster functions Try to use only fast functions.
► Identifying mathematical functions You need identify fast
You can to be selective
about mathematical the time
functions by knowing
► Identifying standard functions complexity associated with
Use standard mathematical
functions, such as square
root, function.a program.
the used in
approaches to compute a
► Declaring local functions as static Using static functions,
result that uses complex
You can sequence and
evaluation is faster
calculations. functions based
mathematical
efficiency is improved.
This enables you to solve a
on execution time.
problem more efficiently.



Examining Branching

Transferring control from one part of code to another is
achieved through branching.
Various techniques can be adopted to make the process of
branching effective and improve the efficiency of code.
Branches can be examined by:
► Removing the else clause Using an else clause with
every if loop leads to
► Using effective case statements You should use effective
ineffective branching.
case statements in such a
► Replacing conditional computations Therefore, youcomputational
You can save should try to
way so the order of options is
remove replacingclause, if
time by the else conditional
defined based on the
possible.
computation with equivalent
frequency of using the
arithmetic expressions.
options.



General Guidelines for Optimization

Some guidelines to follow when writing code are:
Identify optimization areas
Identify the depth of optimization
Identify correct alternatives
Identify what is being asked



Common Misconceptions About Optimization

Some of the common misconceptions about code and their
optimization are:
To think a program doesn’t require optimization because it
appears to be quite fast.
To think that only the optimization performed by the compiler is
enough.
To believe that short code is efficient.
To think a specific solution will be effective without verifying
performance results.
To think optimizing while programming is good practice.



Just a minute

What is unrolling?
What is sentinel value?

Answer:
Breaking compact loops into simpler statements is called
unrolling.
A sentinel value is a value that is placed at the end of a search
range. This assures the termination of the search.



Using Performance Libraries For Applications

Each software developer has an individualistic style of
writing code. It is possible that the written code will not be
efficient.
Sections of software, such as the menu bar, remain
common in most software.
Writing these sections of codes repeatedly makes the
process time-consuming and error-prone.
In such situations, you might prefer to use existing code,
which is called a performance library.
This piece of code is already checked for errors. In addition,
this code attains a high level of optimization over a period.



Using Performance Libraries For Applications (Contd.)

The advantages of using performance libraries are the
following:
Enable programmers to develop code in less time
Offer error free codes
Enables best use of resources
Help in performance improvement
Add to stability of the functioning software application



Identifying the Types of Performance Libraries

Based on these tasks, performance libraries can be
categorized as:
► Engineering and science libraries These libraries are mainly
used in scientific and
► Math libraries These libraries deal with the
engineering applications.
evaluation of complex
► Graphic libraries These libraries help draw
may include
mathematical functions, such
graphics, pie for searching,
functionalitiescharts, graphs,
► Audio/Video libraries as vector and matrixoptimize
These libraries help
and bar diagrams correctly
sorting, and evaluating
calculations.
► Image processing libraries various functions related to
and accurately.
variouslibraries help process
These expressions.
audio-visual data.
images faster. may include
► Other libraries Other libraries
functionalities for performing
various tasks, such as
speech recognition, signal
processing, and
cryptography.



Just a minute

Which libraries may include functionalities for searching,
sorting, and evaluating various expressions?

Answer:
Engineering and science libraries



Summary

In this session, you learned that:
Time complexity is the time taken by the steps of an algorithm
to execute.
Space complexity is the measurement of the space utilized by
the components of an algorithm.
Identifying the data structures used in an algorithm and then
adding them can measure space complexity.
Loops are generally the most time-consuming constructs of a
program.
Some of the techniques to optimize loops are:
Removing unwanted parts of loops
Combining loops
Using Unrolling
Reducing work inside loops



Summary (Contd.)

Using Sentinel Values
Looking at the order of loops
Looking at operators
– Some of the techniques to optimize functions are:
Using faster functions
Identifying mathematical functions
Identifying standard functions
Declaring local functions as static
Transferring control from one part of a code to another is
achieved through branching.



Summary (Contd.)

Some of the techniques to make the process of branching
effective are:
– Removing the else clause
– Using Effective Case Statements
– Replacing Conditional Computation
You should keep the following optimization guidelines in mind
when writing codes:
– Identify optimization areas
– Identify the depth of optimization
– Identify correct alternatives
– Identify what is being asked



Summary (Contd.)

Some of the common misconceptions about optimization are:
– A fast program does not require optimization
– Optimization performed by the compiler is enough
– A short code is efficient
– Optimizing while programming is good practice
– Performance libraries can improve the performance of an
application to a large extent.
– A variety of performance libraries are available for different
purposes, such as mathematical, graphical, and task-based
functions.


02 intel v_tune_session_02

Recommended

Recommended

More Related Content

What's hot

What's hot (12)

Similar to 02 intel v_tune_session_02

Similar to 02 intel v_tune_session_02 (20)

More from Niit Care

More from Niit Care (20)

Recently uploaded

Recently uploaded (20)

02 intel v_tune_session_02

Editor's Notes